Embed Size (px)
Citation preview
1
Two phases of zymogen granule lifetime in mouse
pancreas: ghost-granules linger after exocytosis of
contents
Peter Thorn* and Ian Parker
Department of Neurobiology and Behavior, University of California Irvine, CA
92697-4550, USA
*Department of Pharmacology, University of Cambridge, Tennis Court Road,
Cambridge, CB2 1PD, UK
Running title: Exocytosis of zymogen granule contents
Key words: exocytosis, ghost-granule, acinar, zymogen
Correspondence should be addressed to:
Peter Thorn,
Department of Pharmacology,
University of Cambridge,
Tennis Court Road,
Cambridge,
CB2 1PD,
UK
email: [email protected]
Physiology in Press; published online on January 6, 2005 as 10.1113/jphysiol.2004.077230
2
Summary
Different cell types show widely divergent mechanisms and kinetics of exocytosis.
We investigated these processes in pancreatic acinar cells by using video-rate 2-
photon microscopy to image entry of extracellular dye into individual zymogen
granules undergoing exocytosis. Fluorescence signals display two distinct phases;
an initial peak that then decays over several seconds to a prolonged plateau.
Several observations suggest that the first component reflects the binding of dye to
the granule contents and their subsequent release into the acinar duct. Among
these, the peak/plateau fluorescence ratio differs between different dyes; the initial
fluorescence decay mirrors the loss of granule contents as monitored by DIC
microscopy; and the fall in vesicular fluorescence is accompanied by a rise in
fluorescence in the adjacent duct lumen. We thus propose the use of extracellular
fluorescent probes as a convenient means to monitor the kinetics of loss of
proteinaceous content from secretory granules. In pancreatic acinar cells the
fusion pore remains open much longer than required to ensure secretion of the
granule contents, and instead the persistent empty ‘ghost-granule’ may act as a
conduit to which secondary granules can fuse and release their contents by
compound exocytosis.
3
Introduction
Recent improvements in imaging techniques now allow us to observe the
behaviour of individual secretory vesicles in real time, revealing a wide diversity
in the mechanisms and dynamics of exocytosis and endocytosis among different
cell types (Jahn & Sudhof, 1999; Taraska et al., 2003; Holroyd et al., 2002;
Graham et al., 2002; Kasai 1999). For example, rates of exocytotic loss of vesicle
contents vary widely between milliseconds (Sabatini & Regehr 1996) to tens of
seconds (Taraska et al., 2003; Barg et al., 2002). Moreover, there are striking
differences in the lifetimes for which vesicles persist after fusion, and in the
mechanisms by which they are recycled. In some cells, fusion results in the rapid
collapse of the vesicle and incorporation of the vesicular membrane into the
plasma membrane, from where it is subsequently retrieved by endocytotic
processes (Valtorta et al., 2001; Heuser & Reese 1973). Very different to this, the
fused vesicle remains intact in other cell types and is recaptured during
endocytosis so that it can be refilled and recycled in another round of exocytosis
(Taraska et al., 2003; Holroyd et al., 2002; Ceccarelli et al., 1972).
Here, we have studied the specialized adaptations of exocytosis in exocrine mouse
pancreatic acinar cells that secrete digestive enzymes in response to neuronal and
hormonal stimuli. These enzymes are contained within secretory vesicles
(zymogen granules) and exocytosis occurs exclusively on the apical plasma
membrane adjacent to the acinar duct (Palade, 1975), leading to enzyme release
into the gut. An unusual characteristic of pancreatic acinar cells, shared with some
4
other cell types, is a process of compound exocytosis, whereby secretory granules
deep within the cell release their contents by fusing with more superficial granules
that have already fused to the plasma membrane. Compound exocytosis was first
characterized by electron microscopy (Ichikawa 1965, Palade, 1975; Amsterdam
et al., 1969), and the sequential fusion of chains of vesicles has recently been
visualized by real time optical imaging (Nemoto et al., 2001; Hafez et al., 2003;
Thorn et al., 2004). A striking feature is that vesicles persist for minutes after
fusion, and remain in continuity with the extracellular fluid through an open fusion
pore (Thorn et al., 2004). We sought to determine whether such long post-fusion
lifetimes are required to ensure emptying of the vesicle contents, or whether the
persistence of empty ‘ghost’ vesicles may be an adaptation to support compound
exocytosis.
To address these questions, we employed extracellular fluorescent dyes to monitor
exocytosis by using high-speed 2-photon microscopy to image entry of dye into
fused vesicles (Nguyen et al., 2001; Thorn et al., 2004). Our results show a
characteristic biphasic time-course of the fluorescent signal as the dye enters the
granules through the open fusion pore. The fluorescence signal rises rapidly to an
initial peak and then decays over several seconds to a plateau that often persists for
minutes. Employing a variety of approaches we conclude that the initial
fluorescence peak is due to dye binding to granule contents, and that its decay
reflects the secretion of these contents (together with bound dye) into the duct.
5
The remaining fluorescent signal during the plateau phase then represents the post-
exocytotic structure (‘ghost’) of an aqueous dye-filled granule.
Methods
Cell preparation
Male outbred mice (~25g) were humanely killed and the pancreas removed
following the local procedures and regulations of the University of California,
Irvine and, for experiments conducted in the UK by the United Kingdom Home
Office. Intact lobules and fragments of mouse pancreas were then prepared using
a modified collagenase (CLSPA, Worthington, USA) digestion technique (Thorn
et al. 1993). The original method was adapted to reduce the duration of
collagenase treatment (to ~8 minutes) and to minimize the extent of mechanical
trituration (~30 passes of the tissue through a 1 ml pipette). The resulting
preparation was composed mainly of pancreatic lobules (~1000s of cells) and
fragments (~50-100 cells), which we plated onto coated (Poly-l-lysine (Sigma-
Aldrich, USA)) glass coverslips. Lobules and fragments were prepared and
maintained in a NaCl-rich extracellular medium composed of (mM): NaCl (140),
KCl (5), MgCl2 (2), CaCl2 (1), Glucose (10), HEPES (10); pH set to 7.4 with
NaOH.
Two-photon imaging
6
We used a custom made, video-rate, 2-photon microscope with a 40 X oil
immersion objective (NA 1.35, Olympus) (Nguyen et al., 2001; Miller et al., 2002;
Thorn et al., 2004) providing a resolution (full width at half maximum) of 0.26 µm
lateral and 1.3 µm axial. Images were acquired at video rate (30 frames per
second) and then averaged (typically 12-45 frames, except for the experiments of
Figure 5, which were acquired and analyzed at video-rate), stored and analyzed
with the MetaMorph program (Ver. 6.0, Universal Imaging, USA).
To image exocytotic events we used the following water-soluble, membrane-
impermeant fluorescent dyes at final concentrations in the extracellular media as
indicated (all from Molecular Probes, Eugene, OR): Sulforhodamine B (SRB, 500
µM); Oregon Green (OG, 100 µM); and Alexa-488 (10 µM). All were visualized
by femtosecond excitation at 780 nm, with emitted fluorescence collected at
wavelengths of 550-650 nm (SRB), 510-555 nm (OG) and 510-555 nm (Alexa-
488). In experiments where OG and SRB signals were recorded simultaneously
(Fig. 2C,D) we used narrow band-pass filters (respectively, 515-555 nm and 620-
650 nm), which effectively eliminated bleed-through of OG fluorescence into the
red channel, and reduced bleed-through of SRB fluorescence into the green
channel to <5%.
We stimulated the cells either by direct application of agonist to the bathing
solution or by photolysis of caged carbachol. For the latter, caged carbachol was
7
added to the bathing solution (final concentration 2.5 - 20 µΜ) and carbachol was
liberated by illuminating a spot (~50 µm diameter) centred on the cells of interest
with a high intensity UV light, as previously described (Thorn et al., 2004).
Imaging of fixed cells in the experiments shown in Fig. 3 was performed as above.
Dye accumulation was assessed by comparing a 7.75 x 7.75 µm region in the
apical domain with a region of the same area outside the cells.
Data are expressed as + SEM; statistical significance was calculated using
unpaired Student’s T tests.
Protein binding to dyes
Experiments carried out to determine the behaviour of dyes in the presence of
protein were done using the 2-photon microscope to measure fluorescence from
the centre of 100 µl droplets of dye solution prepared either in extracellular
medium alone, or in medium including 10% bovine albumin (BSA, Sigma-
Aldrich, USA). To determine the amount of dye that bound to protein we loaded
dye/protein solution on a 10 kDa filter (Centricon, Millipore, USA) and separated
the free dye by centrifugation according to the manufacturers protocol. The
fluorescence intensity of the recovered free dye solution was then compared to the
intensity of the dye in protein-containing medium.
8
Results
Imaging single-granule exocytotic events
We imaged lobules and smaller fragments of mouse pancreatic tissue that retained
the typical morphology of the distal portion of intact exocrine glands (Fig. 1A),
where a network of branching ducts terminate at acini comprised of clusters of
acinar cells enclosing an aqueous acinar lumen. Inclusion of a fluorescent probe
(SRB, OG, or Alexa-488) in the extracellular bathing medium labeled the acinar
lumen and the extracellular space between cells, but all these dyes were excluded
from the cell interior (Fig. 1A right). Visualization by 2-photon microscopy
provided a thin optical ‘slice’ within the tissue, and minimized out-of-focus
fluorescence from the large volume of dye in the surrounding solution. In order to
more clearly visualize the behavior of individual granules we studied spontaneous
events (which arise at a frequency of ~0.05 events per cell per minute within a
given optical slice), as well as responses to lower concentrations (~50 nM) of
acetylcholine that evoked only sparse events in each cell. Single events appeared
abruptly as fluorescent spots (diameter ~0.7 µm; Fig. 1E) in the apical pole;
consistent with dye entering single zymogen granules through the open fusion pore
(Fig. 1B, see also movie event.avi in supplementary material). Measurements of
fluorescence recorded from regions of interest centered on these spots displayed a
characteristic time course comprised of a rapid rise to an initial peak, followed by
a decay over several seconds to a sustained plateau level (Fig. 1C). All 3 of the
9
dyes we used gave qualitatively similar fluorescence signals following granule
fusion (i.e. peak followed by plateau).
We first considered whether these phases may arise from changes in
geometry of the granule. For example, if the granule contracted shortly after
opening of the fusion pore, that would reduce the volume of aqueous dye
contained within the granule causing the fluorescence to decline to a lower plateau
level. To test this we plotted the average fluorescence along a line drawn through
the granule centre at the peak and plateau (measured 20s after peak) of events
recorded with SRB (single event Fig. 1D and the mean of n = 8 events Fig. 1E).
No detectable change in granule width was apparent. However, cannot rule out
changes in granule size below our limit of resolution, and any such changes would
be amplified in the fluorescence intensity signal because vesicle volume scales as
the third power of its diameter.
Different dyes give different peak/plateau ratios
We next quantified the fluorescence signals for the different dyes (Fig. 2A). With
SRB the amplitude of the plateau (measured 10 s after the peak) was 16.1 +/- 4.9
% (mean +/- SEM) of that of the initial peak (5 events), whereas the corresponding
value for OG was 34.2 +/- 5.0 % (13 events), and for Alexa-488 61.97 +/- 5.3 %
(10 events). In unpaired Student’s t tests the plateau level for Alexa-488 was
10
significantly different from that for SRB (P<0.01) and OG (P<0.01), the SRB and
OG plateau levels were also significantly different (P<0.05).
Since any change in granule volume would be expected to give similar peak-
plateau relationships for the different dyes these data provide evidence that the
dyes cannot be acting simply as fluorescent reporters of the water-filled volume of
the granule. Instead, other dye-specific factors appear to influence their
fluorescence.
One possibility is that pH-dependent changes in dye fluorescence (Lattanzio,
1990) might account for these observations, since the granule contents are initially
acidic but neutralize following fusion as the granule interior equilibrates with the
extracellular medium (De Lisle & Williams, 1987). However, even though OG
fluorescence is strongly pH sensitive, it is quenched at acid pH and would thus
expected to show a rise - not fall - in fluorescence if the decay of the first
component were to reflect neutralization of the vesicular pH. Moreover, we found
the fluorescence of neither SRB nor Alexa-488 is pH sensitive between pH 3.0 –
7.4 (data not shown).
Dye interaction with granule contents
11
We next considered whether the proteinaceous granule contents (zymogens) might
interact with the dyes to increase their apparent fluorescence. The initial
fluorescence peak might then result as freely-diffusible dye molecules rapidly
enter the granule and bind to its contents, whereas the subsequent decay of
fluorescence reflects the slower loss of content and its associated dye. In that case,
the fluorescence during the plateau phase would reflect simply the amount of
aqueous dye remaining in the granule.
To facilitate further analysis we thus normalized records obtained with different
dyes to the same plateau level. Fig. 2B compares mean fluorescence signals
obtained using SRB (n = 5 granules) and Alexa-488 (n = 9), after normalizing to
the plateau fluorescence, and demonstrates the dramatically greater initial peak
seen with SRB. Moreover, simultaneous records of SRB and OG signals, obtained
using a mixture of both dyes in the bathing solution and dual photomultiplier
detectors at appropriate wavelengths (Fig. 2C,D), showed consistent differences in
peak/plateau ratios. The plateau level of fluorescence measured 10 s after the peak
was 31.7 +/- 3.8 % (mean +/- SEM, n = 8) of the peak amplitude for SRB, as
compared to 55.9 +/- 5.7 % for OG (significant using a Student’s t test, P <0.01).
Thus the differences in peak/plateau ratios obtained with the dyes alone are still
seen when dyes are imaged simultaneously.
The absolute values of the peak/plateau ratios were different when the dyes were
imaged separately (as in Fig. 2A), as compared to when they were imaged
12
simultaneously. It is not clear whether this is due to differences in the optical setup
needed to for simultaneous measurements, or whether it may reflect competition
between dyes for binding sites. The important point is that peak/plateau ratios
differ appreciably between various dyes, and that differences persist even when
two dyes are imaged simultaneously. Thus, the kinetic changes in dye fluorescence
cannot simply reflect volume changes of the granule, but instead the fluorescent
properties of different dyes appear to be differentially affected by the contents of
the granule.
Characteristics of dye binding to proteins
To characterize how the dye behaviours are modified in the presence of proteins,
we measured their fluorescence when equilibrated with a 10% solution of bovine
serum albumen (BSA). With respect to their fluorescence in aqueous solution,
SRB showed only a slight decrease on addition of BSA, whereas both Oregon
Green and Alexa-488 were quenched to <25% (Table 1). To then determine the
amount of dye bound to protein we filtered the BSA-dye solutions through a 10
kDa cut-off filter (BSA has a molecular weight of ~70 kDa), and measured the
fluorescence of the filtrate. All three dyes bound strongly to BSA (Table 1).
Although these data do not replicate the environment of the granule, and give no
information regarding the kinetics of the dyes on the timescale of granule fusion
and filling, they do show that whereas all three dyes bind strongly to proteins, they
13
are differentially quenched. The fluorescence signal measured from vesicles is
thus expected to depend on the combined effects of accumulation and quenching
as dyes bind to the granule contents. Specifically, SRB bound strongly to protein,
but showed little quenching, consistent with the notion that accumulation of dye
by binding to granule contents may explain the large initial fluorescence transient
seen with this dye. In contrast, OG and Alexa-488 binding was accompanied by a
strong quenching, so that these dyes would be expected to show smaller initial
fluorescence transients.
To further test this hypothesis, and determine dye behaviour with the actual
granule contents, we imaged paraformaldehyde fixed, Triton X100 (0.1%) treated
acinar cells that were bathed in SRB (0.5 mM) or OG (100 µM). After a short
period of equilibration, cells were imaged with the 2-photon microscope (Fig. 3).
In both cases the dyes accumulated in the cells raising their fluorescence above
that of the surrounding media. Regions of interest (7.75 x 7.75 µm) in the granular
apical pole region were 3.66 +/- 0.93 (mean +/- SEM, n = 5) times brighter than
the bathing solution surrounding the cell for SRB, and 1.18 +/- 0.03 (n = 4) times
brighter for OG (P<0.05). These data are consistent with differences in the
peak/plateau fluorescence ratios observed following granule fusion in live cells
shown in Fig. 2.
We conclude, therefore, that the initial large peak of SRB fluorescence following
granule fusion arises because dye entering the granule binds to protein contents,
14
resulting in an effective concentration increase. The subsequent decay in
fluorescence then reflects the loss of protein content together with its associated
bound dye, with the final plateau level corresponding to the dye remaining in free
aqueous solution in the ‘empty’ granule.
Loss of granule contents measured by differential interference contrast
microscopy
A key point in the above interpretation is that the decay of the initial transient of
SRB fluorescence reflects the loss of zymogen contents from the granule into the
secretory duct. To verify if this is the case, we independently measured changes in
granule contents by means of differential interference contrast (DIC) microscopy
(Ishihara et al., 2000). We observed increases in light intensity at discrete spots in
the apical pole following stimulation by acetylcholine (Fig. 4A). The mean rise
time of the DIC signal matched well to the falling phase of SRB fluorescence
signals measured separately in other experiments (Fig. 4B). Single exponential fits
to the decay of the SRB fluorescence and rise of the DIC signal gave respective
time constants of 5.6 +/- 0.8 s (n = 13), and 4.37 +/- 0.6 s (n = 7), in agreement
with the idea that both signals reflect the time course of loss of granule contents.
Are granules empty during the plateau phase of the fluorescence signal?
Following the initial fluorescence transient, plateau levels of fluorescence persist
as long as 12 minutes (Thorn et al., 2004), indicating that the granule is in stable
15
equilibrium with the external medium and may have lost most, if not all of its
protein contents. To test this, we used data from the experiments of Fig 2C,D to
measure the ratio of SRB fluorescence to OG fluorescence in vesicles at the peak
and at the plateau, and compared this to the fluorescence ratio in the bathing
solution (measured in an ~ 10 x10 µm region outside the acini). We reasoned that
at the peak, dye binding to granule contents would result in a higher SRB/OG
fluorescence ratio than in the aqueous external medium. However during the
plateau, the ratio inside the granule should be similar to that outside if the granule
were devoid of proteinaceous contents. Measurements showed a peak SRB/OG
fluorescence ratio of 2.4 +/- 0.35 (n = 9) times larger than that outside, whereas
during the plateau the ratio was only 1.2 +/- 0.37 times greater (n = 8). Thus, the
fluorescence during the plateau phase of the fluorescence signal is consistent with
that expected if the granule contains only an aqueous solution of dye after
exocytosing almost all of its cargo.
16
Fusion events are followed by a fluorescence increase in the adjacent lumen
If the initial fluorescence peak represents dye binding to vesicular contents, the
subsequent loss of contents might be associated with a transient increase in
fluorescence in the adjacent lumen as the protein (and the dye bound to it) passes
out of the granule. This was indeed the case. Measurements of SRB fluorescence
from a region of interest located in the acinar lumen closely adjacent to fusing
granules showed a small rise throughout the period when the granule fluorescence
was falling (Fig. 5, see also event2.avi). Quantitative measurements were made by
placing regions of interest (0.625 µm x 0.625 µm) over the acinar lumen adjacent
to granules undergoing exocytosis (Fig. 5Ac). To compare records across
experiments the lumen fluorescence was normalized to the peak fluorescence
changes. This mean normalized fluorescence increased to 45.09 +/- 15.3 % (mean
+/- SEM) (n = 14) when measured 6 s after the peak of the adjacent granule event
(Student’s t test significant, P<0.05 when compared to the normalized fluorescence
3 s before granule fusion) (Fig. 5B, circles). Thus the temporal correlation
between the rise in luminal fluorescence and fall in granule fluorescence is
consistent with exocytotic movement of granule contents and associated bound
dye.
Discussion
17
Our data provide new insights into the mechanisms of secretion in pancreatic
acinar cells. Previous reports have indicated that zymogen granules (secretory
vesicles) persist for long times after fusion with the apical cell membrane (Nemoto
et al., 2001; Thorn et al. 2004). We now show that this lifetime can be divided into
two phases. The first involves the loss of granule contents over several seconds
following opening of the fusion pore. Subsequent endocytosis is delayed, and the
granule persists for a long time (tens of seconds or minutes) with its fusion pore
open during its second phase as a substantially empty ‘ghost’ (Thorn et al., 2004).
The long lifespan of the ghost-granules is, therefore, not needed to ensure
secretion of their contents, but instead is likely to be an adaptation to permit
compound exocytosis by subsequent fusion of secondary granules to the ghost.
Fluorescent dyes report granule content
Our finding of rapid exocytosis of zymogen granule contents is based on
fluorescence measurements of extracellular dye entry into granules following their
fusion. Fluorescence signals show biphasic kinetics; a rapid (<1s) rise to a peak,
followed by a decay over several seconds to a low plateau that persists for many
seconds or minutes. Based on several lines of evidence we conclude that the
biphasic fluorescence kinetics arise from the binding of dye to granule contents
which are subsequently lost into the duct lumen. (1) We did not observe any
measurable change in granule width associated with the decay of the initial
fluorescence transient, suggesting that the decay does not result from a decrease in
18
granule volume. (2) Different dyes showed different peak-plateau relationships,
indicating that the biphasic fluorescence profile is determined, at least in part, by
the specific properties of different dyes in the intra-granule environment. (3) In
agreement with this, the dyes we employed showed differential binding and
fluorescence quenching in the presence of proteins (BSA) (4) Dye accumulation in
the granule-dense apical domain of fixed, permeabilized pancreatic acinar cells
matched that expected from the live-cell imaging experiments. (5) The time-course
of the fluorescence decay from peak to plateau closely mirrored that of differential
interference contrast signals that are presumed to reflect loss of granule contents.
(6) The SRB/OG fluorescence ratio in the granule during the plateau is similar to
that of the bathing medium, as expected if the ghost granule contains only an
aqueous solution of dye. (7) The decay of granule fluorescence was accompanied
by a transient rise in fluorescence in the adjoining duct. This is consistent with the
exit of dye bound to granule contents, although we cannot rule out other
possibilities such as possible post-exocytotic changes in dye-protein binding
characteristics.
Taken together these data support the idea that extracellular dye enters a granule
upon the opening of the fusion pore and rapidly accumulates on the intra-granule
contents. The subsequent decay in fluorescence then occurs due to the loss of
granule contents and the concomitant loss of bound dye. Fluorescence remaining
during the plateau phase most likely represents dye present in free aqueous
solution in an empty ghost vesicle, although we cannot exclude the possibility that
19
a small fraction of the vesicle contents may persist during this phase.
Measurements of vesicular fluorescence thus offer a promising means to monitor
the loss of proteinaceous granule contents: but, conversely, the effects of dye
binding and quenching should be borne in mind as a possible complicating factor
when using extracellular dyes to study the kinetics of vesicle fusion.
Methods for measuring loss of vesicle contents upon exocytosis
A number of different methods have been used to follow exocytotic loss of vesicle
content. These are variously applicable to measuring secretion, and are discussed
below along with our new method.
One method involves the tagging of a known vesicle protein with a fluorescent
protein, so as to image its loss during exocytosis. (Taraska et al., 2003; Barg et al.
2002). A key advantage is that the decrease in fluorescence provides a direct
measure of loss of the fluorescent protein. Disadvantages include the difficulty of
expressing exogenous proteins in acutely isolated cell preparations, the possibility
that over-expressed proteins may not follow the behaviour of native proteins and
the susceptibility of fluorescent proteins to pH and other environmental factors.
An alternative approach utilizes differential interference contrast microscopy of
vesicles as an indirect measure of loss of vesicle contents (Ishihara et al., 2002).
This provides an indication of the time course of loss of vesicle content, but is
20
difficult to quantify, and signals may be contaminated by factors (e.g. pH change)
other than the content loss.
A final method, of which our approach is a novel variant, tracks the loss of vesicle
contents through the use of vital dyes. This was first demonstrated by Angleson et
al., (1999) who found that that the lipophilic probe FM1-43 also stained the
contents of lactotroph granules, which could then be imaged in real time to follow
granule content loss. However, FM1-43 binding to granule contents has not
generally been observed, and in our hands FM1-43 does not stain zymogen
granule contents in pancreatic acinar cells (Thorn et al., 2004: but see Gaisano et
al., 2001). Instead, the approach we describe is based on aqueous protein-binding
dyes that do not have the complication of binding to lipids. This method is
especially amenable to pancreatic acinar cells, where the granules are densely
packed with protein and the duct is likely to contain little protein at rest. It remains
to be determined whether it will be more generally applicable to other cell types
that secrete proteins. For example, Takahashi et al., (2002) present evidence that
fluorescence signals from granules in pancreatic β cells are consistent simply with
dye occupying the aqueous volume of the granule, and not with dye binding to
content.
Time-course of granule content loss
21
By measuring the decay of the initial fluorescence transient, we estimate that fused
zymogen granules lose their contents (primarily peptide enzymes, such as
amylase: MW ~50 kDa) with a time constant of about 6s. This rate is similar to
measurements obtained using GFP-tagged peptides to monitor secretion in other
cell types. For example, loss of EGFP-islet amyloid polypeptide in an insulin
secreting cell line occurs over 1-10s (Barg et al., 2002), and loss of EGFP-
neuropeptide Y from vesicles in PC-12 cells is complete within 10s (Taraska et
al., 2003). In contrast to these measures of peptide release, indirect measures of
vesicular release of non-peptide neurotransmitters at synapses are some 50,000
faster (Sabatini & Regehr, 1997), pointing to an enormous diversity in kinetics of
exocytosis among diverse cell types that likely reflects both functional and
structural differences (Kasai, 1999). In contrast to neuronal signaling, there is no
especial need for rapid secretion of digestive enzymes: indeed, transport along the
duct is likely much slower than emptying from the granule. Moreover, mechanistic
constraints will limit the speed of release from zymogen granules. Different to the
almost instantaneous release of small-molecule neurotransmitters following kiss-
and-collapse fusion, the larger granules in pancreatic acini remain intact and their
protein contents must escape through the restricted dimensions of the intact fusion
pore.
Structural stablization of ghost-granules
Zymogen granules show little or no decrease in volume after exocytosis (Fig. 1).
Because the granule lipid membrane itself has no structural integrity, some other
22
mechanism must structurally stabilize the granule. One candidate is the actin
cytoskeleton, which forms a coat around individual granules (Valentijn et al.,
2000; Nemoto et al. 2004). New evidence that indicates that this F-actin coat is
formed after opening of the fusion pore and may, therefore, be involved in
maintaining the structure of the primary vesicle to prevent its collapse and,
possibly, to facilitate secondary fusion events (Turvey & Thorn, 2004).
Conclusions
In conclusion our data suggest that, after fusion with the plasma membrane, loss of
content occurs relatively rapidly. The subsequent persistence of the empty granule
provides a target for the fusion of secondary granules and a conduit for the loss of
their contents.
Supplemental data
1. A movie sequence of a single exocytotic event recorded using SRB (event.avi).
Images are 7.7 µm wide and the clip is 40 s long.
2. A movie sequence of the event observed in Fig. 5 is shown, recorded using SRB
(event2.avi). During the sequence the subsequent increase in fluorescence in the
duct is seen. Images are 2.56 µm wide and the clip is 62 s long.
23
Acknowledgements
We thank Mike Edwardson for comments on the manuscript. This work was
funded by a Wellcome Trust travel award (between PT and IP), a Medical
Research Council (UK) project grant (G0000214 to PT) and a grant from the NIH
(GM48071 to IP).
References
Angleson JK, Cochilla AJ, Kilic G, Nussinovitch I & Betz WJ (1999). Regulation
of dense core release from neuroendocrine cells revealed by imaging single
exocytic events. Nature Neurosc. 2, 440-446.
Amsterdam A, Ohad I, & Schramm M (1969). Dynamic changes in the
ultrastructure of the acinar cell of the rat parotid gland during the secretory
cycle. J Cell Biol 41, 753-73.
Barg S, Olofsson CS, Schriever-Abeln J, Wendt A, Gebre-Medhin S, Renstrom E
& Rorsman P (2000). Delay between fusion pore opening and peptide
release from large dense-core vesicles in neuroendocrine cells. Neuron. 17,
287-99.
Ceccarelli B, Hurlbut WP & Mauro A (1972). Depletion of vesicles from frog
neuromuscular junctions by prolonged tetanic stimulation. J Cell Biol 54,
30–38.
24
De Lisle RC & Williams JA (1987). Zymogen granule acidity is not required for
stimulated pancreatic protein secretion. Am J Physiol 253, G711-719.
Gaisano HY, Lutz MP, Leser J, Sheu L, Lynch G, Tang L, Tamori Y, Trimble WS
& Salapatek AM (2001). Supramaximal cholecystokinin displaces
Munc18c from the pancreatic acinar basal surface, redirecting apical
exocytosis to the basal membrane. J Clin Invest. 108,1597-611
Giovannucci DR, Yule DI & Stuenkel EL (1998). Optical measurement of
stimulus-evoked membrane dynamics in single pancreatic acinar cells. Am
J Physiol 275, C732-C739.
Graham ME, O’Callaghan DW, McMahon HT & Burgoyne RD (2002). Dynamin-
dependent and dynamin-independent processes contribute to the regulation
of single granule release kinetics and quantal size. Proc Nat Acad Sci 99,
7124-7129.
Hafez I, Stolpe A & Lindau M. (2003). Compound exocytosis and cumulative
fusion in eosinophils. J Biol Chem 278, 44921-8.
Heuser JE & Reese TS (1973). Evidence for recycling of synaptic vesicle
membrane during transmitter release at the frog neuromuscular junction. J
Cell Biol 57, 315–344.
Holroyd P, Lang T, Wenzel D, De Camilli P. & Jahn R (2002). Imaging direct,
dynamin-dependent recapture of fusing secretory granules on plasma
membrane lawns from PC12 cells. Proc Nat Acad Sci 99, 16806-16811.
Ichikawa A (1965). Fine structural changes in response to hormonal stimulation of
the perfused canine pancreas. J. Cell Biol. 24, 369- 385.
25
Ishihara Y, Sakurai T, Kimura T. & Terakawa S (2000). Exocytosis and movement
of zymogen granules observed by VEC-DIC microscopy in the pancreatic
tissue en bloc. Am J Physiol 279, C1177–C1188.
Jahn R & Sudhof TC (1999). Membrane fusion and exocytosis. Ann Rev Biochem
68, 863– 911.
Kasai H (1999). Comparative biology of Ca2+-dependent exocytosis: implications
of kinetic diversity for secretory function. Trends Neurosci 22, 88-93.
Lattanzio FA (1990). The effects of pH and temperature on fluorescent calcium
indicators as determined with Chelex-100 and EDTA buffer systems.
Biochem Biophys Res Comm 171, 102-108.
Miller MJ, Wei SH, Parker I, & Cahalan MD (2002). Two-photon imaging of
lymphocyte motility and antigen response in intact lymph node. Science
296, 1869-1873.
Nemoto T, Kimura R, Ito K, Tachikawa A, Miyashita Y, Iino M & Kasai H
(2001). Sequential-replenishment mechanism of exocytosis in pancreatic
acini. Nature Cell Biol 3, 253-258.
Nemoto T, Kojima T, Oshima A, Bito H & Kasai H (2004). Stabilization of
exocytosis by dynamic F-actin coating of zymogen granules in pancreatic
acini. J Biol Chem 279, 37544-37550
Nguyen QT, Callamaras N, Hsieh C & Parker I (2001). Construction of a two-
photon microscope for video-rate Ca(2+) imaging. Cell Calcium 30, 383-93.
Ohara-Imaizumi M, Nakamich Y, Tanaka T, Katsuta H, Ishida H & Nagamatsu S
(2002). Monitoring of exocytosis and endocytosis of insulin secretory
26
granules in the pancreatic b-cell line MIN6 using pH-sensitive green
fluorescent protein (pHluorin) and confocal laser microscopy) Biochemical
J. 363, 73-80.
Palade G (1975). Intracellular aspects of the process of protein synthesis. Science,
189, 347-368.
Sabatini BL & Regehr WG (1996). Timing of neurotransmission at fast synapses
in the mammalian brain. Nature 384, 170-172.
Takahashi A, Camacho P, Lechleiter J & Herman B (1999). Measurement of
intracellular calcium. Physiological Reviews 79, 1089-1125
Takahashi N, Kishimoto T, Nemoto T, Kadowaki T, & Kasai H (2002). Fusion
pore dynamics and insulin granule exocytosis in the pancreatic islet.
Science 297, 1349-1352
Taraska JW, Perrais D, Ohara-Imaizumi M, Nagamatsu S & Almers W (2003).
Secretory granules are recaptured largely intact after stimulated exocytosis
in cultured endocrine cells. Proc Nat Acad Sci 100, 2070-2075.
Thorn P, Lawrie A M, Smith PM, Gallacher DV & Petersen OH (1993). Local and
global cytosolic Ca2+ oscillations in exocrine cells evoked by agonists and
inositol trisphosphate. Cell 74, 661-8.
Thorn P, Fogarty KE & Parker I (2004). Zymogen granule exocytosis is
characterized by long fusion pore openings and preservation of vesicle
lipid identity. Proc Natl Acad Sci 101, 6774-6779.
27
Tsuboi T, Zhao C, Terakawa S & Rutter GA (2000). Simultaneous evanescent
wave imaging of insulin vesicle membrane and cargo during a single
exocytotic event. Curr Biol 10, 1307-1310.
Turvey MR & Thorn P (2004). Lysine-fixable dye tracing of exocytosis shows F-
actin coating is a step that follows granule fusion in pancreatic acinar cells
Pflugers Archive, Euro J Physiol, 448, 552-555
Valentijn JA, Valentijn K, Pastore LM, & Jamieson JD (2000). Actin coating of
secretory granules during regulated exocytosis correlates with the release
of rab3D. Proc Nat Acad Sci 97,1091-1095
Valtorta F, Meldolesi J & Fesce R (2001). Synaptic vesicles: is kissing a matter of
competence? Trends Cell Biol 11, 324-328.
28
Table 1.
Fluorescence in presence of protein
(% control)
% of dye bound to protein
SRB 88 65
Oregon Green 25 87
Alexa-488 20 83
Measurements of in vitro binding of fluorescent dyes to protein (BSA), and
consequent quenching of fluorescence. See text for details.
29
Figure Legends
Figure 1. Morphology of pancreatic acini, and imaging of individual fusion
events
A, Left image shows a transmitted light image of a typical pancreatic fragment
used for experiments. Right image shows an optical section through the same
fragment imaged using 2-photon microscopy to excite a fluorescent dye (OG) in
the extracellular medium. Individual cells exclude dye and appear negatively
stained, whereas exocrine ducts between cells are clearly visible. B, An example of
a single exocytotic event (indicated by arrow). The sequence of 2-photon images
show the initial fluorescence in an acinar lumen which traverses the image in a left
bottom to right top diagonal (i); the abrupt appearance of a bright fluorescence
spot immediately following granule fusion; and the dimmer fluorescence
persisting in the granule about 30 s later (iii). Note that further events occur in the
left side of the acinar lumen. C, Graph shows corresponding measurements of
fluorescence obtained over time from a region of interest (1 µm x 1 µm) located
over the granule. An initial large peak in fluorescence decays rapidly to a sustained
plateau level. Times at which the images in (B) were captured are indicated. D,
The fluorescence intensity of a line drawn through the granule in (B) at the peak
(circles) and at the plateau (diamonds) shows that although the fluorescence signal
decreases, the width of the granule remains similar (scaled plateau, triangles). E,
Mean fluorescence intensity measured (with SRB as the extracellular dye) along a
line drawn diametrically through individual granules. The mean fluorescence at
30
the peak of the event (circles), and at the plateau (20s after the peak, diamonds) are
shown, together with the plateau fluorescence scaled to match the peak (triangles).
The mean width of the granules (full-width at half maximal intensity) was 0.77 +/-
0.05 µm (n = 8) at the peak and 0.77 +/- 0.13 µm at the plateau (no significant
difference; P = 0.31)
Figure 2. Different fluorescent dyes show qualitatively similar kinetic signals
following fusion events, but display large differences in relative amplitudes of
the peak and plateau components
A, Traces were derived from images like those in Fig. 1B, and represent mean
kinetics of several events imaged with three different dyes (OG, n = 13 : SRB, n =
5 : Alexa-488, n = 10), after normalizing to the initial peak fluorescence. All
records show spontaneous events in unstimulated cells. B, Comparison of the
decay phase of the initial fluorescence transient, after normalizing to the plateau
level. Traces show mean fluorescence recorded in separate experiments with SRB
(thick line, n = 5) and Alexa-488 (dashed line, n = 9). C, Simultaneous traces of
SRB and OG fluorescence during an individual granule event, obtained using dual-
channel imaging with extracellular solution including both dyes. D, Mean records
from 9 events (all spontaneous) obtained using simultaneous, dual-channel
imaging of SRB and OG fluorescence.
31
Figure 3. Two-photon imaging of fixed and permeabilized pancreatic acinar
cells shows differential dye accumulation in secretory granules
Fixed cells were permeabilized with Triton X100 and imaged with a 2-photon
microscope in the presence of either OG (A) or SRB (B) in the bathing solution.
The images were normalized with respect to the brightness of extracellular
fluorescence. Both dyes stained the cells, indicating binding to cell proteins, but
fluorescence in the cell compared to fluorescence in the surrounding media
showed that dye accumulation of SRB was much larger than for OG.
Figure 4. Optical signals measured by DIC microscopy during granule
emptying
A, Thick trace shows transmitted light intensity measured by differential
interference contrast (DIC) microscopy from a region of interest (0.5 x 0.5 µm)
centered on a granule undergoing exocytosis in response to stimulation with caged
carbachol. Dashed line is a single exponential fit to the rising phase. B, Thick trace
shows the mean DIC signal during 7 secretory events, and the thin trace shows the
mean SRB signal (reproduced from Fig. 2D) derived in other experiments. Both
traces are scaled to the same peak magnitude, and are aligned in time to their
onset.
Figure 5. Fluorescence decrease in granules is accompanied by a transient
increase in fluorescence in the adjoining acinar lumen
32
A, Representative example of a single vesicle fusion event followed by a slow rise
in lumenal fluorescence. a, Low magnification image showing 3 acinar cells.
SRB was present in the extracellular solution. b, Enlarged views of the acinar
lumen region marked by the box in (a). Each panel is an average of 37
consecutive frames (captured every 200 ms). The upper row of images show (left
to right) basal fluorescence in the acinar lumen, an exocytotic event, and elevated
fluorescence in the adjacent acinar lumen following the event. Regions of interest
marked on the middle image were used to measure the traces shown in (c). The
lower row of images were obtained after subtracting the initial basal fluorescence.
c, Traces show fluorescence changes measured from regions of interest as
indicated in (b). B, Graph shows changes in SRB fluorescence measured from the
vesicle and adjacent acinar lumen averaged from 16 recordings like that in (A).
Traces show (from top to bottom): granule fluorescence, normalized to peak;
lumenal fluorescence, normalized to its peak; and the raw mean luminal
fluorescence trace.
50 m�
200
0
0
80
time (s)
A.
C.
i
ii
iii
Fig. 1fluore
scence
(a.u
.)
phase fluorescence
1 m�
B.
i ii iii
200
1.60.80
0
distance ( m)�
D.
1
1
2 30
distance ( m)�
fluore
scence
norm
aliz
ed
topeak
peak
plateau (20s)
scaled plateau
E.
Flu
ore
scence
(au)
peak
plateau
scaled plateau
A.
60time (s)
1
0
1
0
0 30time (s)
fluo
resc
en
cen
orm
aliz
ed
top
late
au
B.
25
95
0 30time (s)
fluo
resc
en
ce(a
.u.)
D.
SRBOregon Green
C:
0 30time (s)
flu
ore
sce
nce
(a.u
.)
0
100SRBOregon Green
SRBOregon Green
Alexa-488
SRBAlexa-488
0
Fig. 2
flu
ore
sce
nce
no
rma
lize
dto
pe
ak
Fig. 3
A: OG B: SRB
10 m�
time (s)
time (s)
flu
ore
sce
nce
no
rma
lize
dto
pe
ak
Inte
nsity
(a.u
.)
0
0
0
100
120
1
24
30
DICfitted
A.
B.
DICSRB
Fig. 4
